AI Engineer

Vadim  is a Senior AI/ML Engineer with over six years of experience in designing and deploying production-grade AI and machine learning systems. His core expertise lies in Python-based development, with a strong command of FastAPI, PyTorch, and Hugging Face, used to build scalable microservices and integrate advanced models into production environments. Vadim’s technical background includes deep knowledge of ML pipelines, vector databases (PGVector, cloud solutions), and agentic frameworks such as LangChain, LangGraph, and GraphRAG. He has hands-on experience across AWS (Bedrock, Model Garden), GCP (Vertex AI), and Azure ML, and has worked on projects involving enterprise integrations with systems like Confluence and Jira. He began his career as a Backend Engineer, developing ETL processes and data-driven systems, before moving into machine learning roles. At Imperial Fund, Vadim contributed to projects focused on Named Entity Recognition (NER) for financial documents, combining classical ML with applied data analysis. Currently, he specializes in end-to-end AI pipeline design, covering everything from system architecture to deployment and performance optimization. His work often involves retrieval-augmented generation (RAG) and graph-based reasoning, applying these methods to create intelligent, scalable, and secure AI systems.

Bachelor — Financial University

Computer Science

Senior AI Engineer — May 2023 – Present

• Integrated ML and NLP models into backend services (Hugging Face, TensorFlow, PyTorch) → auto-mated document processing and reduced manual workload by 20%.

• Optimized API performance and database queries → cut response latency by 17% on critical endpoints.

• Built and maintained scalable microservices (FastAPI, gRPC, PostgreSQL, Redis) → ensured reliable data exchange in high-load environments.

• Designed and deployed retrieval-augmented generation (RAG) pipelines → improved search relevance by 15%

Senior Machine Learning Engineer | Imperial Fund — Jun 2020 – Mar 2023

• Developed an automated data ingestion pipeline using Kafka and Airflow, which reduced ETL latency by 25%.

• Designed NLP pipelines (Hugging Face, spaCy) for document parsing and entity extraction, accelerating financial reporting.

• Implemented end-to-end MLOps workflows using MLflow and Kubeflow for model deployment, tracking, and reproducibility

Python Backend Developer | Luxoft — Jan 2019 – May 2020

• Designed RESTful APIs (FastAPI) and ETL workflows (Airflow), reducing manual reporting time by 20%.

• Developed real-time analytics dashboards using scikit-learn and Plotly Dash to provide key operational insights.

• Deployed and scaled backend services using Docker and Kubernetes, ensuring high availability and re- producibility